Topic: Pushpool - Tech Support - page 7. (Read 135255 times)

I ran export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
and that fixed it. But I'll run that too.

just run: sudo apt-get install libevent-dev

On my ubuntu server I got pushpoold compiled and installed but when I try to run it, it says can't locate libevent.so.5 file not found. I installed libevent according to the guide, I also found there is a shortcut for it in /usr/local/lib I'm not too sure what to do. I went into /etc/ld.so.conf and added a line "include /usr/local/lib/*.*" no go. Maybe i'm just being a linux noob 

Thanks mate! I will give it a try and let you know. Cheers for the info.

Wes.

I think it would be one of two things:

1. You're using the wrong bitcoin RPC credentials (I don't think so - you'd be seeing a 401 Authorization Required response) -
2. Pushpool is connecting on the wrong port. That's likely the problem. Make sure the rpc port specific in server.json is the same as the rpcport specified in bitcoin.conf.

Also make sure pushpool isn't listening on any ports that are the same as the bitcoin rpc port.

Check "netstat -au"  and see how many TIME_WAIT handles you have open.  If that number exceeds your ufile limit, that would be why.

However, in my case, I implemented a patch to fix that problem and I'm still seeing it.

I think the RPC multithreading has some issues - I've observed that every once in awhile, it matches the wrong response to an outgoing request.

Hey guys,

Was wondering if you came across this before, my miner tries to connect but doesn't get any work. I can see when miner connects but get the following:

Connecting my miner.

Quote
JSON protocol request:
{"method": "getwork", "params": [], "id":1}

* About to connect() to 127.0.0.1 port 8333 (#0)
*   Trying 127.0.0.1... * TCP_NODELAY set
* connected
* Connected to 127.0.0.1 (127.0.0.1) port 8333 (#0)
* Server auth using Basic with user 'username'
> POST / HTTP/1.1
Authorization: Basic enRqdWg6dGVzdDEyMw==
Host: 127.0.0.1:8333
Accept: */*
Accept-Encoding: deflate, gzip
Content-type: application/json
Content-Length: 45

* Empty reply from server
* Connection #0 to host 127.0.0.1 left intact
[2011-07-14 01:27:16.074094] HTTP request failed: Empty reply from server
JSON protocol request:
{"method": "getwork", "params": [], "id":2}

* Connection #0 seems to be dead!
* Closing connection #0
* About to connect() to 127.0.0.1 port 8333 (#0)
*   Trying 127.0.0.1... * TCP_NODELAY set
* connected
* Connected to 127.0.0.1 (127.0.0.1) port 8333 (#0)
* Server auth using Basic with user 'username'
> POST / HTTP/1.1
Authorization: Basic enRqdWg6dGVzdDEyMw==
Host: 127.0.0.1:8333
Accept: */*
Accept-Encoding: deflate, gzip
Content-type: application/json
Content-Length: 45


Any ideas what you think could be wrong?

Thanks in advance!

Cheers,
Wes.

I've found about every 24 hours or so pushpoold will "lock up" and stop responding to everything. a killall -HUP doesn't help, I have to kill it and restart it and then everything works fine again. It's getting really bad. Does anyone have any tips?

Thanks a lot dude !!!!!

Keeping in mind I don't intend to run either one, if I were to choose one I'd choose simplecoin.  It's a bit more developed, cleaner.  I've looked at the code for both and the sites they produce.

For whatever my opinion on the subject is worth.

my "rpc.target.rewrite" is set to true .
I will post at that thread you gave me thanks .
What do you think is better Mining farm or simplecoin ?
Thanks !

How many workers are connected = up to you to define - the typical approach is you look at all the workers that have reported a share in a given time window, usually 10 minutes.  The number of distinct workers in that list is the count of 'connected' workers.

Workers don't actually "connect" - you can't think of it as a connection staying open.  Data is passed back and forth between pushpool and the miners in a send/receive/end manner, so the concept of 'connected' is not really what you are measuring.  You're just measuring recent active workers.


It's possible pushpoold isn't reporting the shares - what is your  "rpc.target.rewrite" value set to in your Pushpool json config file?  Beyond that, it's a question for Xenland over at Mining Farm tech support - http://forum.bitcoin.org/index.php?topic=18886.0

Hello,
Thanks a lot guys !
How can i see using pushpoold how many workers are connected and the total m/hashes ?

I have another problem with mining farm , i have a connected worker of 300 M/hashes and the front end doesn't recognize it , what should i do? Is it possible that pushpoold is not reporting about it ?

And i am thinking moving to simple coin, as i feel the mining farm has a problem in cronjobs
what do you think ?

Can someone help please ?
Thanks !

It's advisable to have rpcallowip only be the machines you actually want to allow to connect to bitcoind itself.  Thus, you can actually just comment out the whole line.  Bitcoin will always allow connections from the localhost, which is usually where pushpool is running.

If you have pushpool running on a different server, you should include that server's IP in rpcallowip.  You can add as many of those lines as you want, by the way.

Thanks !
I fixed the problem , by making the same rpc port for the 3 files . ( the rpc url , the port in the bitcoin.conf and in blkmond.conf )

I did rpcallowip=*.*.*.*
Can i remove this without any problem letting anyone connecting to the pool ?
Thanks !

It's ok for testing but you're not going to want to do that for production.  For instance I've just finished installing bitcoind and pushpoold as services, and setting up Monit to monitor and restart memcached, bitcoind and pushpoold should they go down.

They'll also start when the machine reboots.  Set and forget.

Nice thing about Monit is it has a web interface you can use for monitoring and service control. Very slick.


These indicate there's a port mismatch between pushpoold and bitcoind.  Make sure that the RPCport in bitcoind, is the one that pushpoold is connecting to.  This took a bit of figuring out for me when I first set these up, the variable names in the config files are not especially enlightening.

I actually have bitcoind open a different port that pushpoold connects to.  Then I have pushpoold expose 8332 to the web for miners.  No particular reason to do so except to maintain the convention of using 8332 for mining.

I reinstalled my OS , so now it is fresh ubuntu 11 without directadmin ( i figured out that i had debian 6 )

and now i could reach this
and only this , but i can't get out to the shell command line , is this ok ? and how can i test my pool ? i think i have a worker test@test

Edit:
I figured that i must connect using port 8344 , but i these errors
HTTP request failed: couldn't conncet to host.
And i get from the miner upstream RPC error.
The port i specified at the bitcoib.conf is 8332 .
Is this the problem ?

Edit2 : i rebooted my server. ran bitcoind
and now
i get initialized ,
but after some time i get
 HTTP request failed: Empty
reply from server.
and the miner cann't connect ,i changed nothing but rebooted :S
What is this ? i will try to play a little
Thanks
Thanks !

I don't know what happend but if those directories don't exist then you will probably want to reinstall MySQL..
Give us the output of that file and I'll take a look into it! Btw we need the mysql-dev or mysql-devel install so it should be in there..

Ah makes perfect sense, thanks.

This is what i see,

I checked in Directadmin and it says that MySQL 5.5.9 is not running , i am trying the running commands , but nothing is available, something deleted it , i am trying to reinstall ...
What do you think ?
Thanks